What is a DS ?

- One in which software or hardware component located at networked computer communicate and coordinate their actions only by passing messages Ex : + Internet + Mobile phone network + Campus network + ome network - !I" is useful for sharing resources # $ resource could take a form of : hardware component %disk & printer & processing& etc' & software component %file& database& ob(ect & etc ' - Ma(or :Concurrency & no global clock & failures # Definition of single tier , dual tiers and 3 tiers architecture : - Single tier : it was form in a huge )main frame* computer # $ central computer ser+ed whole the business community and access +ia dumb terminals# $ll processing tool place on single computer , and therefore one place # $ll resources associated with the computer %disk dri+e & printer &tape & etc' attached to this same computer # + $d+antage : - "imple - Efficient - -ncomplicated + !isad+antage - .he cost of the central machine is +ery expensi+e - Dual tier : In order to pro+ide personal software which ran on the personal computer & a model found where data shared #.his became client/ser+er model # .he client %personal computer' connect to a central ser+er to re0uest data # .his model 0uite easy to implemented # $ client %personal computer' connect to ser+er +ia a .C1/I1 connection & ser+er running on a 2!3M" & such as M" "45 "er+er 6777 # .he application connect to ser+er and re0uest data # .he ser+er side (ust return data # .he problems of this architecture : - .he connections are +ery expensi+e - One can only connect a limited number of users to a ser+er before "45 ser+er spends more time managing connection than processing re0uest - Cost-ineffecti+e : many users use their connection 6-8 9 of the time - 3 tiers architecture : + :ith ad+ent of Internet the day of mainframe were back + Client/ser+er ob+iously fail & personal computer had fail and :indows was on it way out + $ host of )thin client* application where de+eloped + .hus 8 tier architecture were born .hese application run the traditional client/ser+er model but from a web ser+er #.he client only show ;-I and data & but has no part in producing results #

:hat difference between 6 tiers and 8 tiers architecture - 6 tier : 6 layers like client and ser+er & client re0uest a 0uery to ser+er and ser+er responds to the re0uest by fetching data from it # - 8 tiers : 8 layers like client & ser+er and database are in+ol+ed # :hen a client send a re0uest to ser+er where ser+er pass it to database & based on the re0uest !atabase send data back to ser+er and ser+er forward it to client # 3 tiers architecture : - the user interface stored in client - the bulk of business application logic is stored in one or more servers. - the data is stored in a data server. Openness : the openness of a system is the characteristic that determine whether the system can extended the re-implemented in +arious ways # Openness is one of the important features of distributed system # <ey aspect of Openness - "tandard interface and protocol %like internet protocol ' - "upport of heterogeneity %by e0uate middleware & like CO23$' - Open system are characteri=ed by the fact that their key interfaces are published - Open distributed system are based on the pro+ision of a uniform communication mechanism and published interfaces for access to shared resources Heterogeneity : !istributed application are typically heterogeneity : - different hardware : mainframe & 1Cs & workstation & ser+er & etc - different software : ->I? & M" :indows & I3M O"/6 & etc - programming language : @a+a & C++ & $ssembly & etc - implementation by different de+elopers .he solution for this problem is middleware , an additional software layer to mask heterogeneity # Middleware pro+ides a uniform computational model for use by programmers of ser+ers and distributed application Security : - Confidentiality : protection against disclosure to unauthori=ed person - Integrity : 1rotection against alteration and corruption - $+ailability : <eep the resource accessible Scalability : .he system should remain efficient e+en with a signiAcant increase in the number of users and resources connected: - Cost of physical resources should be reasonable # - 1erformance loss with increase number of user and resources should be controlled - "oftware resource should not run out ailure handling : the ability to continue in operation after a fault has occurred !ransparency : - "ccess transparency : local and remote resource can access using the same identical operation # Clients should be unaware of the distribution of the files# .he files could be present on a totally different set of ser+ers which are physically distant apart and a single set of operations should be pro+ided to access these

remote as well as the local files# $pplications written for the local file should be able to be executed e+en for the remote files# .he examples illustrating this property are the Bile system in >etwork Bile "ystem %>B"'& "45 0ueries& and >a+igation of the web# #ocation transparency : enable resources to be accessed without knowledge of their physical and network location # .he location and access transparencies together are sometimes referred as >etwork transparency# .he examples are Bile system in >B" and the pages of the web $oncurrency transparency : -sers and $pplications should be able to access shared data or ob(ects without interference between each other# .his re0uires +ery complex mechanisms in a distributed system& since there exists true concurrency rather than the simulated concurrency of a central system #.he examples are >B"& $utomatic .eller machine %$.M' network# %eplication transparency : .his kind of transparency should be mainly incorporated for the distributed file systems& which replicate the data at two or more sites for more reliability# .he client generally should not be aware that a replicated copy of the data exists# .he clients should also expect operations to return only one set of +alues# .he examples are !istributed !3M" and Mirroring of :eb pages# ailure transparency : enable the concealment of faults& allowing users and application program to complete their task despite the failure of hardware or software components # Examples are !atabase Management "ystems# &obility transparency : allow the mo+ement of resources and clients within a system without affecting the operation of users or programs# 'erformance transparency : allow the system to be reconfigured to impro+e performance as loads +aries Scaling transparency : allows the system and application to expand in scale without change to the system structure or the application algorithms# .he bestdistributed system example implementing this transparency is the :orld :ide :eb# 6 tiers architecture .he most important limitation of the two-tier architecture is that it is not scalable because each client re0uires its own database session 8 tiers architectures The key 3-tier benefit is improved scalability since the application servers can be deployed on many machines.

"calability

performance reliability